A production line running the same LED-cure adhesive, at the same wavelength, on the same lamp for months, can suddenly start producing tacky bonds or soft cores with no obvious change to the process — and the instinct to blame the lamp is usually wrong. Most LED-cure defects trace back to the adhesive’s own formulation chemistry interacting with a process variable, not equipment failure.

Free-Radical and Cationic Chemistry Behave Differently Under the Same Lamp

Not every adhesive marketed as “LED-curable” reacts the same way once the light hits it. Free-radical acrylate systems — the majority of LED-cure adhesives in electronics and general bonding — react almost instantly under exposure but stop reacting the moment the light source is removed, and they’re sensitive to oxygen at the surface. Cationic epoxy systems, used less often but present in some optical and high-temperature LED-cure formulations, react more slowly, keep curing for a period after the light turns off (a useful property for shadowed geometry), and aren’t affected by oxygen inhibition at all. Diagnosing a defect without knowing which chemistry is in the tube leads to the wrong fix nearly every time — a cationic-system “slow cure” complaint is often normal behavior, not a defect.

Diagnosing the Five Defects That Actually Show Up on the Line

• Tacky surface, solid interior: almost always oxygen inhibition on a free-radical system — ambient oxygen scavenges surface free radicals before they can propagate the reaction. A shorter wavelength (365nm tends to overcome surface tack better than 395–405nm), a nitrogen-purge fixture, or switching to a surface-cure-optimized formulation each address this differently, and only one may be practical for a given line.
• Soft interior, cured surface: a dose problem, not an intensity problem — the surface received enough energy to gel while the bulk didn’t, typically because exposure time was cut to hit a cycle-time target rather than validated against the adhesive’s actual depth-of-cure rating.
• Undercured edges on an otherwise good part: a shadow or beam-uniformity issue — light intensity commonly falls off well before the rated working distance, and a part’s own geometry can shade its own edges from an array that isn’t centered correctly.
• Bond that cured fine but delaminates after thermal cycling: rarely a cure defect at all — more often a CTE mismatch between the cured adhesive and substrate expressing itself only after repeated expansion and contraction, a mechanism covered in depth in Incure’s guide to CTE mismatch and adhesive bond failure.
• Bond that never gels at all despite full exposure time: a wavelength-photoinitiator mismatch — every photoinitiator package has a specific absorption peak, and an adhesive formulated for 365nm run under a 405nm-only array may receive plenty of total energy without ever absorbing the wavelength it needs.

Why Adding More Light Isn’t Always the Fix

The instinct when a bond undercures is to increase intensity or dwell time, and sometimes that’s correct — but oxygen inhibition specifically gets worse, not better, past a certain intensity threshold in some acrylate formulations, because the faster surface reaction generates more heat and volatilizes more surface inhibiting oxygen into the still-liquid layer below it. Dose (irradiance multiplied by time) is the right lever for a genuine depth-of-cure shortfall; it is the wrong lever for a surface-tack complaint that’s actually a chemistry-and-atmosphere problem. Reading a defect correctly before adjusting the lamp settings — rather than escalating intensity as a default response — avoids wasting a validation cycle on the wrong variable. Email Us with the defect pattern, adhesive chemistry (free-radical or cationic), and current wavelength if a formulation-level diagnosis would help before the next production run.

Matching Formulation Class to the Actual Joint

Beyond troubleshooting an existing defect, the formulation itself should be chosen for the mechanical demand of the joint rather than defaulting to whatever cured the last project successfully:

• Rigid, high-modulus formulations suit structural bonding where the joint sees direct load and minimal flex — automotive sensor housings and metal-to-metal structural joints typically fall here.
• Flexible, high-elongation grades are the correct choice wherever thermal cycling or vibration will flex the bond line repeatedly over its service life — a rigid formulation in that same application tends to develop fatigue cracks at the interface well before a flexible one would.
• Low-outgassing, optically clear chemistries are non-negotiable for lens bonding and precision optical assembly, since standard formulations can release volatiles that fog nearby optical surfaces over time even after the bond itself appears fully cured.

Frequently Asked Questions

Q: Can a puck test alone confirm a bond is fully cured?
A: A puck test (measuring cured thickness of a sample under the same exposure) confirms depth of cure for that specific geometry, but doesn’t rule out oxygen-inhibited surface tack, which requires a separate surface-tack check.

Q: Does switching from a 395nm to a 405nm array fix most undercure problems?
A: Only if the adhesive’s photoinitiator actually absorbs at the new wavelength — a wavelength change without checking the formulation’s absorption spec is a common, avoidable mistake.

Q: Is a cationic system ever a drop-in replacement for a free-radical one?
A: Rarely without requalification — the two cure through entirely different chemical mechanisms, and a lamp/process combination validated for one won’t necessarily produce the same result with the other.

A defect that looks like a lamp problem is more often a formulation-and-process mismatch, and correctly identifying which of the two is at fault saves a validation cycle that would otherwise chase the wrong variable. Incure’s Uni-Weld and Optik LED-curable lines are formulated with published wavelength and depth-of-cure data specifically so this diagnosis doesn’t have to happen by trial and error on the production floor. For the equipment side of this same relationship — matching a lamp’s actual delivered output to what a formulation needs — see Incure’s guide to UV LED curing. Contact Our Team to review a recurring defect against your current adhesive chemistry and cure parameters.

Visit www.incurelab.com for more information.